9. Reptiles SPECIES SOUTHERN NORTH ISLAND MARLBOROUGH SOUNDS. Brothers Island tuatara Tuatara bones from middens on Mana North Brother Island

Transcription

1 9. Reptiles Designing a future reptile community on Mana Island required information on: what indigenous species are currently present what species are likely to have been present historically what nationally or regionally threatened species require habitats free of introduced mammals to ensure their continued survival, and whether there are other more appropriate sites for their introduction potential conflicts between proposed introductions and resident species and/ or other proposed introductions (including invertebrates and other reptiles), e.g., predation, competition, disease risk, hybridization. Mana island already has nationally important populations of two threatened lizard species (McGregor's skink and goldstripe gecko) and it is essential that future introductions and habitat management do not jeopardise their survival. The list of reptiles recorded from the entire Cook Strait Ecological District is not an appropriate model for guiding restoration of Mana Island's reptile fauna, as many lizard taxa do not occur on both sides of Cook Strait. Of 17 lizard species recorded from Cook Strait Ecological District, only eight (47%) occur on both sides of Cook Strait (Table 9.1). Most notably, Cook Strait is the southern limit of the skink genus Cyclodina, five species of which occur (or occurred) in the Wellington region. TABLE 9.1 REPTILE SPECIES RECORDED FROM COOK STRAIT ECOLOGICAL DISTRICT, WITH A SUMMARY OF PAST AND PRESENT DISTRIBUTION ON BOTH SIDES OF COOK STRAIT. NOTE THAT PACIFIC GECKOS OCCUR NEARBY IN THE WELLINGTON ECOLOGICAL DISTRICT, AND WERE PROBABLY PART OF THE ORIGINAL COOK STRAIT REPTILE FAUNA. OGLE (1989a) RECOMMENDED THAT THE TWO SIDES OF COOK STRAIT BE INCLUDED IN DIFFERENT ECOLOGICAL REGIONS OR DISTRICTS BASED ON DIFFERENCES IN FAUNA EITHER SIDE OF THE STRAIT. SPECIES SOUTHERN NORTH ISLAND MARLBOROUGH SOUNDS Brothers Island tuatara Tuatara bones from middens on Mana North Brother Island Cook Strait tuatara Island (species not determined) Stephens &Trio Islands Copper skink Widespread (including Mana Island) absent Robust skink Middens on Mana Island absent McGregor's skink Mana island absent Ornate skink Kapiti Island and mainland absent Whitaker's skink Pukerua Bay absent Speckled skink Wairarapa Stephens Island 75

2 TABLE 9.1 CONTINUED Spotted skink Matiu & Makaro Islands, rare elsewhere On many islands and mainland Common skink Widespread (including Mana Island) Widespread Brown skink Kapiti & Mana Islands and west coast On many islands and mainland Goldstripe gecko Mana Island absent Duvaucel's gecko Middens on Mana Island Several islands (incl. Brothers &Trios) Forest gecko Kapiti Island and mainland Maud, Motuara and Long islands and mainland Common gecko Widespread (including Mana Island) Widespread "Marlborough mini" gecko Wellington south coast Widespread Striped gecko absent Stephens and Maud Islands Wellington green gecko Kapiti Island and mainland absent Marlborough green gecko absent Stephens,Arapawa & D'Urville Islands and mainland As the reptile fauna of Cook Strait Ecological District is not an appropriate model for restoring the Mana Island reptile community, reptiles recorded from the Cook Strait coast of the southern North Island were used as a starting point for restoration. Reptile species were identified as being candidates for introduction if they met one of the following sets of criteria: (a) (b) species recorded in middens on Mana Island that were likely to have had a resident population, and that are unlikely to have significant impacts on threatened plant or animal species that survived on Mana (3 species). species present in the southern North Island that are declining in the presence of mammalian predators, are likely to be able to establish a self-sustaining population on the island, and are not expected to compromise other conservation values on the island (4 species). Restoration of Mana Island's reptile fauna was considered in relation to potential ecological restoration on Kapiti and Matiu/Somes Islands, to ensure that at least one island population is established for each reptile species that is extinct or declining on the southern North Island mainland. 9.1 CURRENT SITUATION Six reptile species are present on Mana Island: copper skink, McGregor's skink, common skink, brown skink, goldstripe gecko and common gecko. Copper skink, common skink and common gecko are abundant and widespread on Mana Island and elsewhere. Brown skinks were only discovered on Mana Island in 1996, but they are probably common in the island interior; their distribution in the Wellington 76

3 region is disjunct, but they are locally abundant. The Mana Island populations of McGregor's skink and goldstripe gecko are restricted and of national significance. McGregor's skink occurs on only four islands: Motuharakeke (5.8 ha) in the Cavalli Group, Mauitaha (4.5 ha) in the Outer Bream Group, Sail Rock (3.4 ha) near Hen Island and Mana Island (217 ha), and is currently listed as a Category B species for conservation priority (Molloy & Davis 1994). McGregor's skink occur on less than 5 ha of Mana Island, but the population is increasing steadily since mice were eradicated (Newman 1994). The Mana Island population is of considerable significance as the southernmost population (by over 560 km) and the population with the greatest potential for natural expansion. It is also the most accessible population for research and management purposes. Changes in density and distribution have been monitored annually since 1985 on Mana Island (Newman 1994), and there is a coordinated programme to search other parts of the island for further surviving populations. Goldstripe geckos are found in coastal Taranaki between Urenui and Patea, and on Mana Island; the species is currently listed as a Category C for conservation priority (Molloy & Davis 1994). Only four goldstripe geckos were found on Mana Island between their discovery in 1972 and A targetted survey in February 1993 (three years after mouse eradication) located a minimum of 112 goldstripe geckos at 11 sites scattered over a distance of 1.7 km (Whitaker 1993). Goldstripe geckos were found predominantly in areas of flax (9 of 11 sites), a rare vegetation type on the island. The Mana Island goldstripe gecko population is of national significance as the only island population, the only population free of predation by introduced mammals, and it is possibly the single largest surviving population (Whitaker 1993). Mana Island is also the largest protected land area where goldstripe gecko occur. Future introductions of birds and reptiles to Mana Island as part of an ecological restoration programme must not compromise the continued survival and expansion of these two resident threatened reptile species. 9.2 WHAT WAS MANA ISLAND'S ORIGINAL REPTILE FAUNA? Seven reptile species were identified following archaeological excavation on Mana Island (Horwood 1991 and in lit.; Table 9.2). Four of these species are still present on Mana Island, but the three others are no longer present in the southern North Island. These three species (tuatara, robust skink and Duvaucel's gecko) are all likely to have been resident on Mana Island, and should be high priorities for reintroduction. The list of reptile species recorded from midden deposits on Mana Island is unlikely to be complete, as evident by the apparent absence of goldstripe gecko and brown skink remains from the midden (they may be impossible to distinguish from common gecko and common skink respectively). It is possible that any or all of the 17 reptile species recorded from the southern North Island may have been present on Mana, as all but three (speckled skink, Pacific gecko and "Marlborough mini" gecko) are known to have occurred on Kapiti Island, Mana Island and the coast in between. 7 7

4 TABLE 9.2 REPTILE SPECIES RECORDED FROM MIDDEN DEPOSITS ON MANA ISLAND (MICHELLE HORWOOD IN LIT.). MNI = MINIMUM NUMBER OF INDIVIDUALS REPRESENTED BY BONES RECOVERED. BONES IDENTIFIED BY TREVOR WORTHY. SPECIES MNI NEAREST EXTANT POPULATION Tuatara sp. 4 North Brother Island Robust skink 7 Castle Island McGregor's skink 4 Mana Island Ornate/copper skink 3 Mana Island (copper skink) Common skink 2 Mana Island?Duvaucel's gecko I Brothers Islands Common gecko 4 Mana Island 9.3 REPTILE SPECIES OF THE SOUTHERN NORTH ISLAND THAT MAY REQUIRE ISLANDS FREE OF MAMMALIAN PREDATORS Of the 17 reptile species recorded from the southern North Island, only four (copper skink, common skink, common gecko and "Marlborough mini" gecko) remain widespread and abundant on the mainland (Miskelly 1995; Rod Hitchmough pers. comm.). The brown skink is locally abundant at a few sites, while ornate skink, spotted skink, forest gecko and Wellington green gecko are widely distributed, nowhere abundant and probably declining. Whitaker's skink, speckled skink and Pacific gecko are each known from only one or two sites in Wellington Conservancy, and are rare at these sites. The remaining five species are known only from islands within Wellington Conservancy (McGregor's skink, goldstripe gecko) or elsewhere (tuatara, robust skink, Duvaucel's gecko). Islands free of mammalian predators may be required to ensure the long term survival of up to twelve species of reptiles from the southern North Island. Six of these species are already present on Kapiti Island (ornate skink, forest gecko, green gecko), Mana Island (McGregor's skink, goldstripe gecko) or Matiu/Somes Island (spotted skink), although the three species on Kapiti Island are all very rare there, possibly due to predation by Norway rats, kiore and a variety of predatory birds (morepork, kingfisher, long-tailed cuckoo and weka). The remaining six species (tuatara, robust skink, Whitaker's skink, speckled skink, Duvaucel's gecko and Pacific gecko) will have to be introduced, or reintroduced, to one or more of these islands to ensure their survival in the southern North Island. This will not be possible on Kapiti Island until rat eradication is confirmed. 78

5 9.4 RESTORING MANA ISLAND'S REPTILE FAUNA Possible future reptile faunas for Mana, Kapiti and Matiu/Somes Islands are given in Table 9.3, with more detailed discussion for each species below. If all translocations recommended are completed successfully, there will eventually be 13 reptile species on Mana Island (217 ha), 15 species on Kapiti Island (1970 ha) and 11 species on Matiu/Somes Island (26 ha); are these islands able to support such diverse reptile faunas? TABLE 9.3 REPTILES OF THE LOWER NORTH ISLAND, WITH SUGGESTIONS FOR RESTORING THE REPTILE FAUNAS OF MANA, KAPITI AND MATIU/SOMES ISLANDS. POSSIBLE SOURCE POPULATIONS FOR TRANSLOCATIONS ARE GIVEN IN BRACKETS. NOTE THAT SOME MAINLAND LIZARD POPULATIONS ARE NOW AT VERY LOW DENSITIES AND CONSIDERABLE CATCH EFFORT WILL BE REQUIRED TO OBTAIN SUFFICIENT ANIMALS FOR TRANSFER: ORNATE SKINK, WHITAKER'S SKINK, SPECKLED SKINK (WAIRARAPA), FOREST GECKO, PACIFIC GECKO AND WELLINGTON GREEN GECKO. SPECIES STATUS ON MANA I. STATUS ON KAPITI I. STATUS ON MATIU/SOMES 1. Tuatara Reintroduce (Stephens I)?Introduce (Stephens I) Reintroduce (Brothers Is) Copper skink Present Present Present Robust skink Reintroduce Introduce - McGregor's skink Present Introduce (Mana 1.) - Ornate skink - Present (rare) Introduce (Wellington) Whitaker's skink introduce (Pukerua Bay) Introduce (Pukerua Bay) - Speckled skink introduce (Stephens 1.) Introduce (Stephens 1.) Introduce (Wairarapa) Spotted skink Introduce (Matiu) Introduce (Matiu) Present Common skink Present Present Present Brown skink Present Present Introduce (Makaro) Goldstripe gecko Present Introduce (Mana 1.) - Duvaucel's gecko Reintroduce (Brothers Is)?Present - Forest gecko - Present (rare) Introduce (Wellington) Common gecko Present Present Present "Marlborough mini" gecko Pacific gecko - - Introduce (Upper Hutt) Wellington green gecko Introduce (Wellington) Present (rare) Introduce (Wellington) Number of species

6 There are three main factors that are likely to lead to reptiles failing to establish a new population, assuming that sufficient animals are released: lack of suitable habitat, excessive predation, and competition from ecologically similar species. All 17 species are likely to have been present in coastal habitats in the southwest North Island historically, and suitable habitat for all these species is likely to exist on Kapiti Island with its diverse vegetation and extensive rock talus slopes and boulder beaches. Kapiti Island is also sufficiently large for closely related species to be spatially segregated during establishment. However, two species of rats were/are present on Kapiti (eradication attempted in 1996) along with a guild of predatory birds. Even if rats were successfully eradicated, it will be at least the turn of the century before eradication is confirmed and further checks are made for rare lizard species that may have survived rat predation in numbers too low to detect at present. It is too early to assume that it will be possible to restore a diverse reptile fauna to Kapiti Island. Both Mana and Matiu/Somes Islands are free of rodents, have few resident predatory birds, and are predominantly covered in rank grass and seral shrubland. However, Mana Island is considerably larger than Matiu, and has larger areas of talus slopes and boulder beaches to provide suitable microhabitats for reptiles while the forest cover on the island becomes established. Of the reptile species listed in Table 9.3, almost all can occur and thrive in coastal shrublands, grassland and/or seral forest (Table 9.4). Within the Wellington region, possibly only the forest gecko requires established forest. All the larger Cyclodina skinks are primarily forest dwellers, but can survive in more open habitats provided there are sufficient retreats such as rock crevices, talus slopes, seabird burrows and dense ground cover. Both Mana Island and Matiu/Somes Island are undoubtedly large enough to hold the diverse reptile faunas suggested for them. Middle Island in the Mercury Island group is half the size of Matiu and supports a reptile fauna comprised of 11 species (tuatara, 4 Cyclodina skinks, 3 Oligosoma skinks and 3 Hoplodactylus geckos;towns et al. 1990). 8 0

7 TABLE 9.4 SIZE, HABIT AND HABITAT OF THE SEVENTEEN REPTILE SPECIES RECORDED FROM THE SOUTHERN NORTH ISLAND. SVL IS MAXIMUM SNOUT-TO- VENT LENGTH (IN MM) RECORDED FOR COOK STRAIT SPECIMENS WHERE POSSIBLE (NOT KNOWN FOR ROBUST SKINK, PACIFIC GECKO AND FOREST GECKO). "COVER" REFERS TO ROCKS, LOGS ETC. LYING ON THE GROUND. NOTE THAT ALL LARGER CYCLODINA SKINKS REQUIRE HABITATS WITH DEEP RETREATS SUCH AS BURROWS, BOULDER BANKS, SCREES, LOGS OR DEEP LITTER. DATA FROM ROBB 1980, TOWNS 1992a, WHITAKER 1993 AND PERS. COMM., CREE 1994, NEWMAN 1994, ROD HITCHMOUGH PERS. COMM. AND MISKELLY PERS. OBS.). SPECIES SVL HABIT HABITAT Tuatara 213 Nocturnal, ground-dwelling Forest, scrub/seabird burrows Copper skink 57 Diurnal/crepuscular, ground-dwelling Open sites with cover/scrub Ornate skink 72 Crepuscular, ground-dwelling Forest/scree/scrub with cover Whitaker's skink 100 Nocturnal, ground-dwelling Forest/scrub/scree/seabird burrows McGregor's skink 114 Nocturnal/crepuscular, ground-dwelling Boulder beach/scrub/seabird burrows Robust skink 120 Nocturnal, ground-dwelling Forest/scree/seabird burrows Common skink 63 Diurnal, ground-dwelling Dry open sites with/without cover Brown skink 66 Diurnal, ground-dwelling Moist sites with cover/scrub/forest Spotted skink 95 Diurnal, ground-dwelling Open sites with cover/scrub/burrows Speckled skink 115 Diurnal, ground-dwelling Forest/scrub with cover/burrows Common gecko 65 Nocturnal, ground-dwelling/arboreal All habitats "Marlborough mini" gecko 50 Nocturnal, ground-dwelling/arboreal Coastal screes, cliffs and boulder beaches Goldstripe gecko 75 Nocturnal/diurnal, arboreal Flax/scrub Pacific gecko 85 Nocturnal, arboreal/ground-dwelling All habitats Forest gecko 87 Nocturnal/diurnal, arboreal Forest/scrub Duvaucel's gecko 116 Nocturnal, ground-dwelling/arboreal Cliffs/scrub/forest Wellington green gecko 95 Diurnal, arboreal Scrub/forest 8 1

8 Figure 9.1 Recorded sympatry in reptiles of the southern North Island, based on records at specific sites for islands, or records from the same 10,000 metre grid square on the mainland. Data mainly from Amphibian and Reptile Distribution Scheme database administered by Department of Conservation. Note that all 120 potential species pair combinations have been recorded within one Ecological District. = reported sympatry for respective species pair, O = no reported sympatry. Some records are from subfossil remains (Worthy 1987 & 1991) therefore temporal sympatry is unproven in these cases. Species not intended for introduction to Mana Island are shaded to highlight that only eight new sympatry pairings will be created during restoration of the Mana Island reptile fauna (see text for discussion). While all the reptile species under discussion have been broadly sympatric in the past, not all species pairings have been recorded from the same site (Figure 9.1). This raises the issue that some species may be ecologically incompatable with each other, and that inappropriate introductions may lead to one or more species becoming locally extinct through competitive interactions. However, there are few data to support this possibility, as most species pairs that are known to have similar ecological requirements (Table 9.4) are currently sympatric at some sites, e.g., copper/ornate skinks, ornate/mcgregor's skinks, Whitaker's/robust skinks, common/ brown skinks, spotted/speckled skinks, common/goldstripe geckos, goldstripe/ 82

9 Pacific geckos, common/pacific geckos, Pacific/forest geckos. It is possible that ecological interactions may be more complex than revealed by species-pair comparisons, for example there are no known sites in Taranaki where goldstripe, Pacific and common geckos occur together (Tony Whitaker pers. comm.). The seven reptile reintroductions/introductions proposed for Mana Island will create eight species pairings that have not previously been recorded at a single site (Figure 9.1). In all eight cases the two species involved have very different ecological requirements; in seven cases one species is diurnal and the other is nocturnal, and in the remaining case (Whitaker's skink/goldstripe gecko) one species is ground-dwelling and the other arboreal. There is a possibility that the diurnal speckled skink could occupy the same habitats and retreats as the large, nocturnal or crepuscular Cyclodina species, and so liberation sites for these species should be widely separated. A pen trial to assess competition for retreats between McGregor's skink and speckled skink should also be attempted. Perhaps the most contentious reptile introduction issue for Mana Island is whether Whitaker's skink should be introduced, given the presence of the larger, aggressive McGregor's skink, and plans to reintroduce robust skinks. All three species are large (maximum weights 20 g for Whitaker's skink, 40 g for McGregor's skink and 70 g for robust skink), all are nocturnal and have similar physiological requirements (Cree & Daugherty 1991). Whitaker's and robust skinks are naturally sympatric on two islands (Middle Island and Castle Island; Table 9.5), but McGregor's skink no longer occurs at any site where either of the two other species occurs. Historically all three species occurred throughout the North Island, and their bones have been found together in one cave deposit (Tapuwae Weka Cave, Waitomo; Worthy 1987). McGregor's skink and robust skink were sympatric at several sites including Mana Island. Large Cyclodina skinks are extremely vulnerable to mammalian predators, and the current distribution of robust, McGregor's and Whitaker's skinks probably reflects sites at which they were able to escape predation either through the absence of mammals (eight islands), or unusual features of their habitat that provided protection from predators (Pukerua Bay and Mana Island). The eight northern islands with these large Cyclodina skinks are all tiny, ranging in size from 1 ha to 13 ha, and so it is hardly surprising that these three large Cyclodinas have failed to survive at any one site due to the effects of chance, stochastic events and/or competitive interactions. Perhaps what is most surprising is that on the largest island (Middle Island, 13 ha) an additional large Cyclodina (marbled skink, c.15 g; D.R. Towns pers. comm.) has survived, this is the only location where three large species of Cyclodina are currently sympatric, although the same three species have recently been introduced to nearby Korapuki, Red Mercury and Stanley Islands. Mana Island is sufficiently large that all three large Cyclodina species could be widely separated during establishment of robust and Whitaker's skinks, and it is unlikely that all three would occur at any one site for many decades. By the time any of the three populations expands sufficiently for their ranges to overlap all three species should be sufficiently abundant and widespread for each species to have a competitive advantage in its preferred habitats. 8 3

10 TABLE 9.5 SYMPATRY IN THE THREE LARGE CYCIODINA SKINKS KNOWN FROM THE SOUTHERN NORTH ISLAND. ALL SUBFOSSIL LOCATIONS ARE FROM WORTHY (1987 & 1991), EXCEPT ROBUST SKINK FROM MANA ISLAND (SEE TOWNS 1992b). McGregor's skink only Robust skink only Whitaker's skink only McGregor's skink Robust skink Motuharakeke (5.8 ha; Cavalli group) Mauitaha (4.5 ha; Outer Bream group) Sail Rock (3.4 ha; near Hen Island) Matapia Island (2 ha; Northland) 2 sites, Northland (subfossil) Moturoa Island (9.5 ha; Northland) Green Island (4 ha; Mercury group) 12 sites,waitomo (subfossil) Coonoor region, southern Hawkes Bay (subfossil) 6 sites, Martinborough (subfossil) 2 sites,waitomo (subfossil) Pukerua Bay Tokerau Beach, Northland (subfossil) Otangaroa Station Cave, Northland (subfossil) 3 sites,waitomo (subfossil) Mana Island (robust skink only from midden deposits) McGregor's skink - Whitaker's skink Robust skink Whitaker's skink McGregor's skink Robust skink Whitaker's skink Motutapu Island (subfossil) Middle Island (13 ha; Mercury group) Both introduced to Korapuki Island (18 ha), Red Mercury Island (225 ha) and Stanley Island (100 ha) in the Mercury Island group Castle Island (3 ha; off Hot Water Beach) Opening Day Cave,Waitomo (subfossil) Tapuae Weka Cave,Waitomo (subfossil) 9.5 RECREATING A REPTILE COMMUNITY FOR MANA ISLAND The following annotated list includes six species that are currently present on Mana, seven species that should be introduced (or reintroduced) to the island, and four species that are not considered appropriate for introduction within the next 20 years even though present (or formerly present) in the southern North Island. Species that should not be introduced in the next 20 years (unless new information is obtained on habitat requirements and/or impacts on other species) are listed in square brackets. Species to be introduced (or reintroduced) to the island are listed in bold lettering. Possible release sites are shown in Figure

11 Figure 9.2 Map of Mana Island showing possible liberation sites for reptiles, and the current distribution of McGregor's skink. 8 5

12 Cook Strait tuatara [ Brothers Island tuatara] Brothers Island tuatara has a category A national conservation priority (Davis & Molloy 1994) and medium conservancy priority (Department of Conservation 1996); Cook Strait tuatara has a category B national conservation priority and medium conservancy priority. Both species of tuatara are present in the Cook Strait region. Tuatara bones (species not determined) have been recovered from middens on Mana Island. The Tuatara Recovery Group has recommended that Cook Strait tuatara be released on Mana Island. There are many current proposals to introduce or reintroduce tuatara to islands (e.g., Cuvier, Red Mercury, Stanley, Moutohora, Titi and Matiu/Somes Islands). These programmes are likely to provide information on release and monitoring methods that will be used to guide establishment on Mana Island. Tuatara are predators on other reptiles and large invertebrates, and so care should be taken to ensure that liberations of tuatara on Mana Island do not jeopardise liberations of other threatened species or populations of resident threatened species on the island. For this reason, release sites for tuatara should not be on the shore platform (habitat for McGregor's skink on northeast coast and possible release site for Whitaker's skink on southwest coast) or near Forest Valley (proposed release site for robust skink and Duvaucel's gecko). The intrinsically low reproductive rate of tuatara (Cree 1994) and the low number of animals likely to be released should ensure that there is ample time for all other reptile species to become established on Mana Island before tuatara are sufficiently numerous to limit population expansion of other species (see Towns 1994). One issue that has not yet been resolved is whether tuatara are dependent on the presence of burrowing seabirds. All of the islands where tuatara currently occur support populations of burrowing petrels. However, there is increasing evidence that tuatara were abundant at some mainland sites in the absence of petrels (Worthy & Holdaway 1994) and that the current distribution of tuatara and petrels has been determined by the absence of mammalian predators rather than an obligate dependence of tuatara on petrels. Liberation of tuatara on Matiu/Somes and Mana Islands will be opportunities to test the relationship between tuatara and burrowing petrels. Select a release site that meets the ecological and physiological requirements of tuatara, is distant from proposed released sites for other threatened reptiles, and away from public tracks (in consultation with Tuatara Recovery Group). The most likely release site is a north-facing amphitheatre part way down the cliff east of the trig (Figure 9.2), where there is a small colony of sooty shearwaters. Another possible release site is near the main sooty shearwater colony on a clifftop at the southwest of the island. Prepare the release site by digging artificial burrows, and any other measures recommended by the Tuatara Recovery Group. Plan for release in 1999 or soon after. Continue with attempts to attract burrowing seabirds to Mana Island. 8 6

13 Copper skink Low conservancy priority (Department of Conservation 1996). Common throughout Mana Island. May increase with increasing scrub cover and more extensive forest margins, but unlikely to remain abundant in dark forest interior. None required. Some monitoring occurs as by-catch during McGregor's skink monitoring. Robust skink Category B (Davis & Molloy 1994); medium-high conservancy priority (Department of Conservation 1996). Known from midden deposits on Mana Island. Robust skinks are thought to be primarily forest-dwelling, but their current distribution is confined to scrub and low forest with dense seabird colonies on small rodent-free islands. The release site chosen has high invertebrate prey densities and could easily be modified to provide humid microclimates required by robust skinks (Cree & Daugherty 1991), however Forest Valley is unlikely to have high densities of burrowing seabirds for many decades. The current distribution and habitat use by large Cyclodina skinks has been dictated by the presence or absence of rodent predators. Release of robust skinks into Forest Valley on Mana Island would provide information on habitat use, including whether robust skinks are dependent on the presence of burrowing seabirds. Prepare Forest Valley as a release site by digging artificial burrows, placing logs and slabs of untreated wood on the forest floor, and erecting leaf litter traps. Liaise with Cyclodina Recovery Group to identify source population and timing for release (preferably 1999 or soon after). Conduct pitfall trapping at proposed release site for at least one summer season prior to release to determine whether McGregor's skinks are present. Continue revegetation programme and attempts to attract burrowing seabirds to Mana Island. McGregor's skink Category B (Davis & Molloy 1994); high conservancy priority (Department of Conservation 1996). Confined to less than 5 ha of the north-eastern shore platform. Although the population is increasing and expanding in range since mice were eradicated, the intrinsically low reproductive rate of large Cyclodina skinks (Cree 1994) and the sedentary behaviour of McGregor's skinks on Mana Island (Newman 1994) indicate that it will be many decades, if not centuries, before all suitable habitats on the island are colonised. While currently confined to a shore platform of cobbles and boulders, it is expected that McGregor's skinks will spread through forested areas as revegetation progresses, especially if burrowing seabirds are present. An existing survey and monitoring programme seeks to ascertain whether other pockets of McGregor's skinks still survive on Mana Island. Although there is one old record (1972) from the southwest coast, there is no intention to reintroduce 8 7

14 McGregor's skinks to this site, as it is close to a possible release site for Whitaker's skink. A pittrapping survey was conducted at this site during and , and no McGregor's skinks were found. Continue monitoring population dynamics of the known population, and southward expansion onto Shingle Point. Conduct pitfall trapping from McGregor's Rock north to the mouth of Tauhinu Valley. Continue revegetation programme and attempts to establish populations of burrowing petrels on Mana Island. [ Ornate skink] Low conservancy priority (Department of Conservation 1996). Although widely distributed in the Wellington region, ornate skinks are rare and difficult to locate on the mainland. They are present on Kapiti Island, but are currently very rare there presumably due to predation by rats and weka. Ornate skinks occur on several northern offshore islands, but are not sympatric with Whitaker's skink at any site. As ornate skinks are only slightly smaller than Whitaker's skink and are likely to inhabit similar habitats, ornate skinks are not recommended for introduction to Mana Island at this stage. Future observations on Kapiti Island (assuming that rats are eradicated and Whitaker's skinks introduced) may establish whether ornate and Whitaker's skinks can coexist, and indicate whether it is feasible to create a lizard community with five species of Cyclodina on Kapiti and/or Mana Islands. Do not introduce ornate skinks to Mana Island unless there is new evidence that five species of Cyclodina (copper, robust, McGregor's, ornate and Whitaker's skinks) can coexist. Whitaker's skink Category B (Davis & Molloy 1994); high conservancy priority (Department of Conservation 1996). The only remaining population of Whitaker's skink in the Wellington region occurs in less than one hectare of habitat at Pukerua Bay (Towns 1992a). The Cyclodina Skink Recovery Group has identified the establishment of an island population of Whitaker's skink sourced from Pukerua Bay as a high priority (Towns 1992b). Neither of the two other islands in the Wellington region are suitable for the introduction of Whitaker's skink in the near future, as there may still be two species of rat plus weka on Kapiti Island, and Matiu/Somes Island does not have extensive areas of rock scree, boulder beach or petrel colonies. The issue of whether to introduce Whitaker's skinks to Mana Island given the presence of McGregor's skinks is discussed above; this section is written assuming that introduction of Whitaker's skinks to Mana Island is appropriate. At Pukerua Bay, Whitaker's skinks occur at a north-facing site at the toe of a scree slope covered with Muehlenbeckia and Coprosma (Towns 1992a). There are many similar sites on Mana Island, although those with a north-east or north aspect are within or close to the known range of McGregor's skinks on the island. The most promising site for liberating Whitaker's skinks on Mana Island is at the southern end of the western bay (Figure 9.2) where they would be separated from known populations of McGregor's skink by 3 km of shore platform. This site has a north- 88

15 west aspect, is densely vegetated, has reasonable densities of nesting penguins and gulls, and receives sunshine from at least 1030 hrs even in mid-winter. Whitaker's skinks have a very low intrinsic reproductive rate and appear to be more ecologically specialised than their congeners (Towns 1994). In reviewing translocation methodology for Whitaker's skinks, Towns (1994) recommended that they be released as early as possible in the restoration programme, or at sites well away from species that are better colonisers or predators. On islands larger than 100 ha, Towns recommended simultaneous release of all reptile species at widely separated sites. Survey the southern end of the western bay on Mana Island to identify a site with similar microhabitats to the Pukerua Bay Whitaker's skink colony, but with a variety of different habitats in close proximity (Towns 1994). Liaise with Cyclodina Skink Recovery Group to determine capture and transfer methodology and timing. Speckled skink Medium-high conservancy priority (Department of Conservation 1996). Speckled skinks have an enigmatic distribution, occurring on Moutohora Island (Bay of Plenty), Stephens Island and at a few sites in central and southern North island and north-west South Island. There are two records from the Wairarapa (Miskelly 1995). The animals on Stephens Island are considerably larger than those recorded from other sites (Hardy 1977, Robb 1980) but no genetic differences are apparent (Daugherty et al. 1994). Introduction to Mana Island would establish an island population in the southern North Island and provide a second population of the large animals from Stephens Island. If a Wairarapa population is located, establishment on Matiu/Somes Island should be investigated (Table 9.3). Speckled skinks are diurnal, and favour damper sites with dappled sunlight compared to the similarly sized spotted skink. Both species occur together on Stephens Island and near Nelson Lakes. On Stephens island speckled skinks are often found in seabird burrows under forest, i.e., in similar habitats and sites to those where large Cyclodina skinks occur on northern islands. There are no known sites where speckled skinks are currently sympatric with robust, McGregor's or Whitaker's skinks (although there is one subfossil site where robust and speckled skinks occur together; Worthy 1991). Large Cyclodina skinks are all nocturnal or crepuscular, and so would be active at different times of day than speckled skinks. However, it is not known how speckled skinks and these other species would interact if there were few retreats available, and pen trials should be undertaken to determine this. The release site chosen for speckled skinks is far from those chosen for robust and Whitaker's skinks to reduce the risk of competition for retreats. The spread of both speckled skink and McGregor's skink should be monitored so that any competitive interactions can be assessed when they eventually come into contact. The release site chosen (Figure 9.2) is on the edge of established plantings and old macrocarpa trees at the mouth of Weta Valley, distant from other skink release sites. 8 9

16 Conduct cage trials to assess competition for refuges between speckled and McGregor's skinks. Prepare release site by ensuring that ample loose cover is present (e.g., rocks and logs). Continue monitoring of the southward spread of McGregor's skink, and study interactions with speckled skinks if/when they come into contact. Spotted skink Medium-high conservancy priority (Department of Conservation 1996). Spotted skinks are abundant on Matiu/Somes, Mokopuna and Makaro Islands in Wellington Harbour. Although not recorded from Mana Island, there are old records from Titahi Bay and Plimmerton (Miskelly 1995). Spotted skinks are now rare or absent from mainland sites west of the axial ranges in the southern North Island. Spotted skinks are diurnal and prefer sunny sites with abundant cover. It is anticipated that spotted skinks will eventually spread throughout the rank grasslands and around the shore platform on Mana Island. The area of habitat suitable for spotted skinks will decline as revegetation progresses, but extensive areas of habitat will remain around the shore, on the cliffs and on the plateau. The other diurnal skinks present/proposed for release on Mana Island are common skink, brown skink and speckled skink, all of which are sympatric with spotted skinks on 150 ha Stephens Island (East et al. 1995). The release site chosen (Figure 9.2) is distant from other reptile release sites and the resident McGregor's skink population, and is similar to sites where spotted skinks are abundant on Matiu/Somes Island. Fifty adult spotted skinks from Matiu/Somes Island were released at a site on the south-eastern shore platform in February Monitor population expansion by pitfall trapping and hand-searching. Common skink Low conservancy priority (Department of Conservation 1996). Extremely abundant throughout grassland, open shrubland and around the shoreline on Mana Island. Absent from the forest remnant and the most established plantings. Likely to decline as forest cover becomes more prevalent, but large areas of suitable habitat will remain around the shore platform, clifftops and around forest margins. None required. Some monitoring occurs as by-catch during McGregor's skink monitoring. Brown skink Low conservancy priority (Department of Conservation 1996). Brown skinks were only recognised as being present on Mana Island in 1996, but there have since been at least ten records spread over a large area of the island's interior. It is presumed that a combination of habitat modification (grazing) and mouse predation reduced 90

17 brown skinks to such low levels that they were not detected during lizard survey work before Brown skinks are likely to increase in numbers and range now that there are large areas of suitable habitat and no mammalian predators. None required. Some monitoring may occur as by-catch during other skink monitoring programmes. Goldstripe gecko Category C (Davis & Molloy 1994); medium conservancy priority (Department of Conservation 1996). Although widespread and locally abundant on Mana Island, there are still areas of apparently suitable habitat, including recent plantings, that have not yet been colonised by goldstripe geckos. On Maria Island goldstripe geckos are most numerous in flax, and so corridors of flax could be used to ensure that goldstripe geckos can colonise all areas of shrubland and forest, and some areas of flax should be retained permanently. The national importance of the Mana Island goldstripe gecko population is such that monitoring by encounter rates during spotlighting should be conducted at five-yearly intervals (Whitaker 1993). This monitoring could be designed to assess the spread of goldstripe geckos into new habitats. Plant flax around wetlands and water storage ponds, and along all valley floors where active revegetation is planned. Ensure that corridors of flax link all areas of suitable habitat for goldstripe geckos. Monitor distribution and abundance of goldstripe geckos on Mana Island at five-yearly intervals. Duvaucel's gecko Not mentioned in Wellington Conservancy CMS (Department of Conservation 1996). Recorded from midden deposits on Mana Island. Duvaucel's gecko is New Zealand's largest extant lizard, and is now confined to islands off the north-east coast of the North Island and in Cook Strait. In the Cook Strait region, Duvaucel's geckos are present on the Brothers Islands, Trio Islands and Sentinel Rock, and possibly Stephens and Kapiti Islands (unconfirmed records). Duvaucel's geckos are nocturnal and use a wide range of habitats, including foraging on flax nectar. It is anticipated that Duvaucel's geckos will eventually spread through forest, scrub and cliff habitats on Mana Island from the preferred release site in Forest Valley. There are no sites where Duvaucel's and goldstripe geckos are currently sympatric, and so it is difficult to predict how the two species will interact [this is the subject of a current MSc study by Halema Flannagan, Massey University]. However, both species must have coexisted on Mana Island in the past, and so it is unlikely that reintroduction of Duvaucels's gecko will have a major impact on the expanding goldstripe gecko population. Goldstripe geckos are present in low numbers in Forest Valley (Whitaker 1993), but this is over 500 m from the largest goldstripe gecko population near the houses. 9 1

18 Assess potential for competition between Duvaucel's geckos and goldstripe geckos. Plan for release in 1998 or soon after. Monitor spread of Duvaucel's gecko and study habitat use compared to goldstripe and common geckos. [ Forest gecko] Medium conservancy priority (Department of Conservation 1996). Forest geckos are widely distributed on the New Zealand mainland and a few offshore islands, but are very difficult to locate on the mainland. Within the Cook Strait region, forest geckos have been recorded from Kapiti, Long, Motuara and Maud islands, but were very rare on Kapiti, Motuara and Long Islands in the presence of rats. Forest geckos occur only on islands with established forest, although the very similar "Westland" gecko and Hoplodactylus nebulosus (both included with forest gecko in H. granulatus until recently; Daugherty et al. 1994) both occur on very small islands with low scrub cover. Forest geckos may require mature trees to ensure sufficient cavities and loose bark to hide in. The seral forest being established on Mana Island may not provide sufficient retreats for forest geckos in the presence of goldstripe, common and Duvaucel's geckos. Coexistence of these four species should be examined on Kapiti island if rats are eradicated and goldstripe geckos introduced. Forest geckos could be considered for introduction to Mana Island in the long term if mature forest is established and if there is little risk of competition with goldstripe geckos. Do not introduce forest geckos to Mana Island. Reassess when mature forest is established and if goldstripe geckos are proven to be able to establish on Kapiti Island in the presence of forest and common geckos. Common gecko Low conservancy priority (Department of Conservation 1996). Extremely abundant in most habitats on island. May increase further as grassland converted to forest. None required. Some monitoring occurs as by-catch during McGregor's skink monitoring. [ Pacific gecko] Currently low conservancy priority (Department of Conservation 1996), however, this may need revision (Raewyn Empson pers. comm.). Pacific geckos are widely distributed in the North island and occur on many of the northern offshore islands (Pickard & Towns 1988). However, Pacific geckos are very rare in the southern North Island, with only three records south of Palmerston North, all in the vicinity of Upper Hutt (Miskelly 1995). Pacific geckos are not nationally threatened, but attempts should be made to establish an island population in the Wellington region using local stock. As there is uncertainty over whether Pacific gecko, goldstripe gecko and common gecko can all coexist at one site (see above), Matiu/Somes Island (where goldstripe geckos are absent) is recommended as a release site for Pacific 9 2

19 geckos from the Wellington region, with the potential to eventually establish a second population on Kapiti Island. Pacific geckos should not be introduced to Mana Island unless there is clear evidence that they can coexist sympatrically with both common and goldstripe geckos. Wellington green gecko Medium-high conservancy priority (Department of Conservation 1996). Green geckos are mainly found in seral forest and scrub habitats, and have been found on only ten offshore islands to date: Great Barrier (27761 ha), Little Barrier (3083 ha), Waiheke (9333 ha), Kapiti (1970 ha), Stephens (150 ha), D'Urville (16782 ha), Arapawa (7785 ha), Adele (88 ha), Green (81 ha) and Codfish (1396 ha). This preference for large islands may be because large land areas are necessary to ensure that there are always some areas of seral forest present (D.R. Towns pers. comm.). However, the apparent preference of green geckos for seral habitats may be an artefact of search effort, as the canopy of mature forest is very difficult to search; green geckos do occur in mature forest as well (A.H. Whitaker pers. comm.). Green geckos are widespread in the Wellington region, but there were few reports in the ten years to 1995 (Miskelly 1995). The decline in green geckos on the New Zealand mainland is thought to be due to predation by introduced mammals, but there is little direct evidence for this. On Stephens Island (where green geckos occur in the absence of rodents), green geckos are exceptionally abundant. The revegetation programme on Mana Island is creating extensive areas of seral forest that will provide large areas of suitable habitat for green geckos. As the forest matures there will be an opportunity to determine habitat preferences of green geckos in the absence of mammalian predators. A healthy and expanding Wellington green gecko population on Mana Island could be used as a source population for other restoration programmes in the Wellington region, e.g., Karori Reservoir. It will be very difficult to obtain sufficient Wellington green geckos from a single site to establish a population on Mana Island. It is envisaged that animals will have to be sourced from a wide area (following a public request for recent sightings). Seek recent reports of green geckos from Wellington region through local media. Trickle release Wellington green geckos into established plantings as animals become available, beginning in

20 9.6 SUMMARY OF ACTIONS REQUIRED TO RESTORE A DIVERSE REPTILE COMMUNITY REPRESENTATIVE OF THE SOUTHERN NORTH ISLAND ON MANA ISLAND Select/confirm release sites Tuatara Robust skink McGregor's skink Whitaker's skink Speckled skink Spotted skink Duvaucel's gecko Wellington green gecko 2. Survey for presence of McGregor's skink Robust skink release site Mouth of Tauhinu Valley 3. Prepare habitat 4. Monitor Tuatara (dig burrows) Robust skink (dig burrows, provide decaying wood, create litter traps) Speckled skink (provide cover) Spotted skink (provide cover) Goldstripe gecko (plant flax corridors) Continue revegetation (provide habitat for geckos and Cyclodina skinks) Attract burrowing seabirds (keystone species to support diverse and abundant reptile fauna) McGregor's skink (annual monitoring of population growth and spread) Goldstripe gecko (5-yearly monitoring of distribution) Establishment and expansion of all reptile species introduced to Mana Island 5. Liaise with recovery groups Tuatara Cyclodina skinks 6. Liaise with Nelson/Marlborough Conservancy to confirm availability of animals for translocation to Mana Island, and to co-ordinate transfers with island restoration programmes in the Marlborough Sounds Cook Strait tuatara Speckled skink Duvaucel's gecko

21 7. Liaise with iwi All translocations 8. Liaise with captive breeders of reptiles Robust skink Whitaker's skink (to hold Pukerua Bay animals until sufficient caught for transfer) Wellington green gecko 9. Identify source populations Robust skink Wellington green gecko 10. Translocations a. As soon as possible between 1998 & 2005 Cook Strait tuatara Robust skink Spotted skink Wellington green gecko Whitaker's skink b. Following further research on interactions with resident species Speckled skink Duvaucel's gecko 11. Research required to finalise reptile community structure on Mana Island Sympatry and niche overlap in Cyclodina skinks (C. aenea, C. alani, C. macgregori, C. ornata, C. wbitakeri) Interactions between speckled skink and Cyclodina skinks Sympatry and niche overlap in Hoplodactylus geckos (H. chrysosireticus, H. duvaucelii, H. granulatus, H. maculatus, H. pacificus) 9 5

22 10. Other vertebrates The Wellington region would have originally supported at least three species of bats (Daniels 1990), four species of leiopelmid frog (Worthy 1987b) and about 19 species of freshwater fish (McDowall 1990). Of these, four species (greater short-tailed bat, the frogs Leiopelma markhami and L. waitomoensis, and the grayling) are now extinct, and two further frogs ("Hamilton's" type frog and Hochstetter's frog) are locally extinct. Two species of bat are thought to be still present, with long-tailed bats on Kapiti Island and in the Tararua and Rimutaka Ranges, and occasional reports of lesser short-tailed bats from the Tararua Ranges. Both species of bats are high priorities for recovery programmes (Molloy 1995). Five species of freshwater fish that survive in the southern North Island are considered threatened (Molloy & Davis 1994): short-jawed kokopu (Category A), giant kokopu and brown mudfish (Category B), and koaro and banded kokopu (Category C) CURRENT SITUATION No bats or amphibians have been recorded from Mana Island (though note reference to a mythical giant frog "moka-mokai a Maru-te-whare-aitu" on Mana Island in Best 1923). The only freshwater fish recorded are shortfinned eels POTENTIAL OF MANA ISLAND AS HABITAT FOR NATIVE BATS, FROGS AND FRESHWATER FISH Suitable habitat for bats will not be present on Mana Island for many decades, as bats require many old hollow trees or trees with loose bark as roost sites. In the long term long-tailed bats, and possibly short-tailed bats, should be introduced to Mana Island, but it would be unrealistic to attempt introductions of bats to Mana Island within the time frame of other restorative actions outlined in this restoration plan. Two species of native frogs have survived in the Sounds-Wellington Ecological Region: Hamilton's frog on Stephen's Island, and Maud Island frog (Bell et al. 1998). The two species cannot be distinguished osteologically (Trevor Worthy pers. comm.) and either species may be represented by bones from the Wairarapa (Worthy 1987b). Both frogs are terrestrial, occurring in rock piles and under logs or deep, damp litter on the forest floor. Mana Island has only one tiny area of rock tumble in the island interior (in Kaikomako Valley) and the forest interior is likely to be too dry to support leiopehnid frogs. While it may be possible to engineer small areas of rock piles to provide habitat for frogs on Mana Island, the effort could not be justified given the extensive areas of apparently suitable habitat on nearby Kapiti Island. If leiopelmid frogs are to be restored to the Wellington region, the top priority must be Kapiti Island following rat eradication. Mana Island has very limited potential as habitat for freshwater fish, as the streams stop flowing in summer, and all three stream outlets reach the sea by flowing 97